Hacker News Reader

Comment by Armisael16 | Hacker News Reader

> They could even do a stickshift EV if they wanted to, though most car companies seem to not think it's worth the complexity/maintenance, and I get the impression that electric motors are hard on transmissions.

Electric motors have a wider power band, so there’s less need for gear shifting. They could get more efficiency with multiple gears, but it’s be a smaller benefit than in an ICE vehicle.

One of the big early problems was that the amount of instant torque that an electric vehicle puts out just shreds a gearbox. You could limit the initial torque in software, but that’d kill the joy of the EV. I’m sure someone has managed to work this by now.

On the other hand, different axles don’t have to use the same gearing ratio. That’s how dual motor Teslas work - the front wheels are geared for higher speeds, and more power is shifted towards those as you go faster. No gear-shifting required.

View on HN

>> You could limit the initial torque in software, but that’d kill the joy of the EV.

All electric cars are torque-limited. They simply have to be. A static electric motor, of the type needed for higher speeds, given full voltage would not only snap bits of the drivetrain but rip apart its own mounts. So they are all limited to various degrees.

The limiting factor on low-speed power for EVs isn't the motor. It's the strength of the drivetrain and ability of the battery system to provide the necessary current/voltage without cooking something.

Is this still true with the use of SynRM motors (as deployed by Tesla)?

All electric motors boil down to magnets pulling and pushing on each other to make movement. Whether the torque protection is via software, circuitry, or mechanical wizardry doesn't really matter. They all need it somewhere to prevent static magnetic forces from pulling everything else apart.

>All electric cars are torque-limited. They simply have to be. A static electric motor, of the type needed for higher speeds, given full voltage would not only snap bits of the drivetrain but rip apart its own mounts. So they are all limited to various degrees.

Electric motors are far, far, more forgiving to drive-train components than sporty vehicles with manuals and medium duty trucks of equivalent power levels because there is no clutch to dump. Sure it's not quite as forgiving as an automatic transmission but it's close.

No. Electric motors are potential torque beasts. That is why they, ALL of them in cars, are mated to motor controllers. The controller may be integrated into the motor assembly so as to give the impression that the motor is tame, but the reality is that every electric motor has the potential for damaging/dangerous levels of torque.

https://en.wikipedia.org/wiki/Motor_controller

"A motor controller is a device or group of devices that can coordinate in a predetermined manner the performance of an electric motor. [...] regulating or limiting the torque, and protecting against overloads..."

Well wouldn't the controller also be used to distribute the inverted power to the poles of the stator in sync with the rotation of the hub? I am pretty sure the controller could be there whether or not it was specifically modulating the torque to be lower.

Motor controllers are necessary with AC motors to supply the 3-phase AC that the motors need in sync with the motor's rotation. (DC motor controllers are simpler, but I don't think DC motors are used in any modern commercially-produced EV.) This generally requires some amount of software.

The motor generally has current and voltage and RPM and temperature limits, and the motor controller keeps everything within those limits. But the motor controller isn't there just for that reason, it's there because the motor couldn't function without it.

DC motors are different; those you could just literally replace your gas pedal with an on/off button and wire it straight from the battery to the motor. Not that that's a good idea.

What? Electric cars have far better low-speed torque than ICEs. How do you think Teslas can do 0-60 so fast?

I don't think you understand the GP comment.

All electric cars are torque-limited certainly != electric cars do not have good low-speed torque.

It is precisely because electric motors are capable of such high torque that they have to be torque limited.

I'm not sure of the power curve for an electric vehicle, but if you add gearing, you introduce losses due to gearbox and you increase complexity.

It's likely that it is not worth it in most applications. Only cars that used gearboxes in EV space (two speed) are usually intended to be able to hit 200mph so they need extremely wide rpm range from their electric motors. I know Rimac used one and Taycan (not surprising since Rimac is partly owned by Porsche).

In Rimac case, they used the gearbox to improve 0-60 times but IIRC they are losing the gearbox for their new car.

Hybrids often use gearboxes. My wife's Prius was essentially an extremely heavy 70 HP equivalent EV.

0 to 10 it was a very fastest car although the acceleration curve flattened out fast and it was between hard and impossible to have enough traction to max it out.

I don't hear much about prius transmissions being destroyed.

Something interesting to think about, I wonder with demographic trends etc if 0 to 10 is a more useful measure of daily driver performance and experience than 0 to 60. My wife's car does 0 to 10 at every stop sign, and technically my car is faster at 0 to 60 but I only did that twice a day at most for the commute, for sportiness reasons I might prefer the worlds fastest 0 to 10 car over a theoretically faster 0 to 60 car.

>something interesting to think about, I wonder with demographic trends etc if 0 to 10 is a more useful measure of daily driver performance and experience than 0 to 60. My wife's car does 0 to 10 at every stop sign, and technically my car is faster at 0 to 60 but I only did that twice a day at most for the commute, for sportiness reasons I might prefer the worlds fastest 0 to 10 car over a theoretically faster 0 to 60 car.

Depends on how long the average on-ramp where you live is.

30-75 measured in distance instead of time would be a good measure where I live.

Being able to gracefully merge no matter how bad of a situation the idiot in front of me leaves behind would be far more beneficial for day to day use than off the line speed.

Prius has CVT, not a "gearbox".

Hybrid is a completely different type of vehicle. Since they utilize an internal combustion engine, they suffer from the same power band issues that purely ICE cars have. This means that a hybrid does have a normal gearbox.

I'm no expert on Priuses but my quick Google showed that Prius has the electric motors "after" the gearbox. This means that electric motor can assist turning of the wheel and that it can regenerate energy but it is not using the gearbox.

One solution to this would be how diesel electric locomotives work. Internal combustion engine generates electricity that is then stored in the batteries and used to power the vehicle. These type of "hybrids" are not popular although I believe there is at least one available where it is used as a range extender.

Rimac already removed it form their vehicle unless I am very misinformed. The Tesla Roadster also doesn't have it. Basically almost non of the EV coming out have it.

I assume Prosche just had the expertise for and it made it easier for them to hit the top level speeds.

> On the other hand, different axles don’t have to use the same gearing ratio. That’s how dual motor Teslas work - the front wheels are geared for higher speeds, and more power is shifted towards those as you go faster. No gear-shifting required.

Yep. Alternatively, the Porsche Taycan has an 2-speed transmission gearbox (cf. https://www.caranddriver.com/news/a28903274/porsche-taycan-t...). I don't think gearboxes are going to become mainstream in EVs, with the exception of performance vehicles like this one.

Do the Tesla’s still have differentials then? It would be cool to eliminate those two but then you’d need a motor for each wheel I guess.

They do for all the models currently on the road.

In the future the Roadster and the Cybertruck both have 3 motors. The rear wheels both have one each, so presumably they can avoid the differential there.

Rivian has an independent 4-motor setup. It's a compelling case of extra parts allowing for simplicity.

It also allows for the ever-entertaining tank turn: https://www.youtube.com/watch?v=yzwM8KE2L3I

Differentials still have a place. They allow wheels to be perfectly balanced when negotiating corners. That can be done with independent electric motors, but never perfectly. Lot of little factors have to be considered in the software. For instance, as a car corners weight shifts, this causes the diameters of pneumatic tires to change slightly. A mechanical differential handles this tiny effect seamlessly. Independent motors on each wheel would have to sense, calculate and react to this via software. That is just one effect. There are also uneven road surfaces to deal with. Any slight error would result in one wheel dragging or trying to accelerate, resulting in instability or at least chewing through tire rubber.

Differentials are cheap and reliable, most going for the lifetime of a car without any issue or even real servicing. They are likely the cheaper option than running multiple motors.

Good point. What about simply powering only the outer wheel? It could look at the steering position to decide when to do that?

So long as you are happy giving up 1/2 of your traction, that would only solve corners if the steering wheel is moving. Say your right wheel hits a bump. Even with zero steering input that right wheel is traveling slightly further than the left wheel. (Up and over the bump is longer than the strait line driven by the left wheel.) There also also times when the steering wheel moves but the car remains strait (crosswinds). The further one looks down this rabbit hole, the more appealing a simple mechanical differential becomes.

I don't expect stickshift EVs to be common either, but still I think it's surprising that no one has made a non-luxury-car EV that isn't direct drive.

Manual transmissions seem like a pretty mature technology at this point. For any car company that already makes a manual transmission car it would be easy to do. They might change the design a little (maybe 3 speeds instead of 5, and use sturdier gears and different ratios). Clearly you don't need a transmission in an EV, and leaving it out is simpler and less maintenance, but still... if it can be done and it makes for a (in some ways) better or more fun car, then it's a shame no one is doing it yet.

I'm currently working on converting a Mazda RX-8 to electric, keeping the transmission and clutch. It'll be interesting to see what it's like to drive. IIRC, the motor (Netgain Hyper-9 144 volt dual-ended version) has a max RPM of 8000, which is close to the 9000 rpm redline of the rotary it replaces. The motor is a lot less powerful than what you'd find in a typical Tesla (which I attribute mostly to it being designed to run at a lower voltage), but having a gearshift should make up for at least some of the difference.

I have a friend who dropped a turbocharged engine out of an old 80s BMW into an RX-8. He ended up having to cut a hole in the firewall and completely redesigning the rack and pinion because the old rack would have gone through the middle of the larger BMW engine. He kept the original gearbox and was always worried the low revving high torque BMW engine was going to murder it. :P

Thankfully I don't have that problem. The Netgain motor is pretty big as motors go and it'll be close to the power steering unit, but still I think I've got enough room there.

Most of the work so far has been making battery boxes. I've got one under each rear seat where the gas tank used to be, one in the front down low where the radiator used to be, a smaller one above it, and another that goes over the motor (which raises the CG more than I'd like, but the other batteries that basically as low as they can be probably offset it). Weight balance should be about the same.

I'll have to do some welding on the back seats, because they're too low for the batteries to fit. Sacrificing about four inches of foam will make for an uncomfortable seat, but it should at least look normal.

How do you plan to do heating? Even if it doesn't get cold where you live, but you'll still need heat (and maybe A/C) to avoid fogging the windshield when it's humid. An electric heater, powered by the same batteries?

I haven't figured that out yet. Probably an electric heater of some sort.

Yeah, I'd have thought a DC motor plus a two-speed transmission would be a no-brainer for the budget end of the market (at least until high power AC drives come down a bit) but maybe not.

What batteries are you using for the RX-8?

DC motors are great if you want the most power with the least price, but for a daily driver AC has a few advantages. AC motors are usually zero maintenance because they don't have brushes to wear out, and it's a lot easier to support regen (which I've heard actually makes a big difference to driving experience; people expect the car to slow down when they take their foot off the gas, and it's disconcerting for it to just keep rolling forward like a freewheeling bicycle).

AC motors and motor controllers are generally more expensive than, say, a Warp9 and a Zilla, but the price difference is gradually dropping. Probably the price difference between AC and DC for a major manufacturer is even smaller than what's available retail to hobbyists doing conversions. And I think more and more conversions are being done with repurposed Tesla and Leaf drive units.

A small, cheap AC motor attached to a 2 or 3 speed transmission would be a lot of fun to drive, I think. No need for a reverse gear even. Maybe not even a clutch or flywheel, if you can match the motor speed with the transmission electronically.

My batteries are about 450 pounds of lithium iron phosphate cells. That's generally what's recommended for conversions. Used Tesla packs have much better energy density, but they're more dangerous and need liquid cooling and are awkwardly shaped. I couldn't think of a good place to put them in the RX-8 where they'd be safe.

In Europe, manual transmissions are generally preferred because they're cheaper, not because they're more fun to drive.

Also European cars tended to have smaller engines, which weren't a great combination with the relative inefficiencies of torque converter style automatic transmissions.

However, newer dual clutch style transmissions work pretty well with small engines - I have a car with a 1l engine and DCT and it is a delight to drive.

Manual transmissions are preferred because on a continent where 1L engines drag around 4000lb cars (because displacement taxes) being able to shift when you want (as opposed to just having to live with whatever gear the OEM engineers who have emissions to consider think you should be in) translates into a much improved driving experience.

They're marginally cheaper but the real key is that they let the consumer side step engineering decisions that were made against the consumers interests.

I thought the cars did that for you by detecting emissions tests and switching to 'emissions mode' just for the test?

> I don't think gearboxes are going to become mainstream in EVs, with the exception of performance vehicles like this one.

I don't think transmissions will be the norm in performance vehicles either, since Tesla Roadster gen 2 will not have it, and Rimac seems to have moved away from it for Concept Two as well.

Teslas solution is to optimize the front and rear motors for different speeds.

There is also an experimental motor that can change "gears" by changing how it drives the windings in the motors. I don't remember the name right now.

Point is there are a few solutions that are simpler and lighter than adding a two-speed gearbox, with relatively little compromise.

I wouldn't be surprised if at some point some manufacturer made a car with optional "stick shift" - with a stick that's just some switches and the effect on driving behavior/torque/... entirely done in software. (Right now electric is hot and new, so nobody is doing much to make an electric car pretend to be something it's not, but that could switch around when electric is the standard)

There is a car exactly like this....in China. It's used for driver training, and I suspect it will start being popular elsewhere. In all of EU if you pass your licence on an automatic(which nearly all electric cars are) you can't drive a manual, you have to pass your licence in a manual car. And that's why that car was created, to allow people to pass their test in a manual vehicle and have a licence that allows driving them, even though the vehicle is electric.

https://jalopnik.com/this-chinese-electric-car-designed-for-...

"The various inputs of clutch and shifter work with software in the car that cause it to behave like a manual, ICE car, with shift points, simulated engine sounds, and it can even “stall” if you let off the clutch too quickly."

Part of me wants to create a fake gearbox and clutch kit for an automatic car that does absolutely nothing, but looks convincing to driving examiners. You can just change "gears" when you feel like it without any actual effect and the clutch is just a dumb pedal.

> and the clutch is just a dumb pedal

Thus completely negating training on a manual?

Learning how to effectively use the clutch pedal to me was like, 90% of learning to drive.

Accelerating to a highway would be a dead giveaway to an examiner. Those who learn to use the clutch well enough in the mandatory 20 hours to do it as smooth as automatic gear are few and far between. And obviously they are the ones who wouldn't need to fake it anyway.

Yeah, it'd be like when a TV character "plays" a guitar with dubbed music without moving their left hand.

This is a complete aside, but I was extremely impressed by the NPCs in the Cyberpunk 2077 video game, where every tune they play has correct movements on the guitar for each chord.

I would think the examiner can also tell because the car does change acceleration, and the engine noise changes.

That said, as an American who prefers a manual, I'm not sure why Europe cares so much. It is a skill to learn, but it isn't that hard to learn: I've taught a few people. They have some rough starts (and avoid hills for a while), but the need to shift isn't something that affects how they drive once they get off my street.

At least here in Poland you can only take the exam in the car provided by the testing centre, you can't have the exam in your own car(with some exceptions like when you need to have a car adjusted to your disabilities)

I believe that's the case all over Europe. That the US apparently or supposedly allows taking a car driver's license exam in your private car is.. extremely surprising.

It's not in UK(I realize it's not part of the EU any more, but even when it was it was the same). You take the exam in your own car or a car provided by the training school. The exam centres don't have cars.

>your own car or a car provided by the training school. The exam centres don't have cars.

Same in Switzerland.

>That the US apparently or supposedly allows taking a car driver's license exam in your private car is.. extremely surprising.

At last here in Indiana, you have to provide your own car to take the driving test in.

US drivers’ licensing is much less rigorous than Europe’s as I understand it. In Arkansas, where I live, I got a “learner’s permit” at 14 having passed a short written test. I got my full license at 16 after passing the road test, which consisted of driving for perhaps ten minutes with a cop in the passenger’s seat. I had to drive around the parking lot, navigate and park between traffic cones, then drive the testing route. The testing route started in an industrial part of town with very little traffic and consisted of driving about a mile to a residential neighborhood, circling the block, and coming back to the testing center. I remember being surprised that I didn’t have to demonstrate parallel parking or even driving on a highway.

I think I “studied” for the written test by skimming the handout brochure at school during lunch, and had no issues. The driving test was a joke; I’m very confident my 7-year-old could pass it today based only on her having driven utility vehicles (e.g., a Kubota RTV) around our family’s property and having watched us drive from the back seat.

Apparently Arkansas revised the licensing system in 2009, and now there is an “Intermediate” license for people under 18. That license is restricted from driving from 11pm to 4am (with exceptions), prohibited from using a cell phone while driving (with exceptions), and all passengers must be wearing a seat belt. If you’re married, in the military, have graduated high school (or equivalent) or have been legally emancipated you can get your full license at the age of 16 years and six months.

This discussion had me curious about how the very different licensing regimes might impact outcomes, so I looked up some statistics on motor vehicle fatalities.

    US (2018):
    * 12.4 per 100k residents
    * 14.2 per 100k vehicles

    Europe (2016)
    * 9.3 per 100k residents
    * 19 per 100k vehicles

    Poland (2019)
    * 7.7 per 100k residents
    * 13.5 per 100k vehicles

Honestly, I don’t think there’s a substantial impact. The US has 838 vehicles per 1k residents compared to Poland’s 719. I strongly suspect that US drivers drive substantially more miles per year on average, which would more than account for the per capita numbers.

Source:

https://en.wikipedia.org/wiki/List_of_countries_by_traffic-r...

WRT fatalities its not apples to apples comparison as USA highways are incredibly expensive with concrete barriers between oncoming traffic and in Ireland the highways between cities are literally two lanes no barriers much like USA farm roads yet with interstate highway levels of traffic.

Meanwhile the cars involved in the accidents are often unimaginably small and lightweight in EU. Rolling a typical Irish car off the road must have an outcome much like rolling a golf cart would, certainly fatal to the occupants, whereas rolling a typical USA SUV into the ditch will ruin the trim and probably the glass but the occupants should be unharmed if they're wearing seatbelts.

I don't think you realize how Americans drive on undivided highways and rural roads nor how strong the cabins of all highway capable vehicles for sale in the US and Europe are.

1) You're comparing 6 million people country with 300 million people country.

2) I've looked M4 and M7 in Ireland on Google Maps and they have two lanes each side with separation.

3) There is plenty highways in Europe with concrete barriers and few lanes each way.

> 1) ...300 million people

The urban density matters more than the country's population

In my state they roll a lot of safety inspection type stuff into the test.

There are no standards for "how to engage the hazard blinkers" across manufacturers so the testing process verifies that at least one time in their driving experience all drivers turned the hazard blinkers on in their private car, and maybe they helped their kids or grandkids figure it out later in life.

I've driven several manual transmission cars and its always so awkward for the first couple minutes as there is no standard "feel". EU test inspectors must provide enormous amounts of leeway during testing. Every time I've driven a new manual car the very first time I let the clutch out is always at least somewhat unexpected. So the USA testers get to use smoothness as a proxy for experience. If you let the clutch out like a guy who's never done it before, you are probably lying about the many hours of required yet otherwise unverifiable pre-testing driving experience.

So in summary it enforces safety education and proves some level of experience.

> I've driven several manual transmission cars and its always so awkward for the first couple minutes as there is no standard "feel". EU test inspectors must provide enormous amounts of leeway during testing. Every time I've driven a new manual car the very first time I let the clutch out is always at least somewhat unexpected.

That's true, but in my experience (France), the car being driven during the exam is one of your driving school's cars, which you are probably familiar with, since a number of driving lessons are mandated before the exam.

Plus, all these cars feature dual controls as a safety measure, not only for pedals, but for blinkers, honk, etc. If the inspector touches their break pedal, you automatically fail the exam.

>EU test inspectors must provide enormous amounts of leeway during testing.

In my experience, no. The difference is that pass rates are at 20-30%.

Exactly. In Poland I did my training in diesel toyta yaris, but the testing centre only had petrol ones. Meaning I obviously stalled the car as soon as I started moving and got the first strike for stalling the car. I had to be super duper careful since I wasn't familiar with the clutch and general feel of a petrol car vs a diesel.

Why is that surprising. Presumably it's representative of the kind of car you'll be driving.

The training school car (which is the one I took my exam in) has a double set of pedals and additional mirrors so that the examiner can take over in case you do a horrible mistake. Which does happen quite a lot - some 20-30% of students fail the driving exam, and sometimes for nearly causing an accident (the examiner intervenes).

I took my test in the UK (which was part of the EU at the time) and I used my own car.

Not the case in Belgium. The test center has no cars, you take the exam in your own car or the car of the driving school.

What would be the issue? It's the same car you practiced in.

I do not think this is the case anymore. I have seen driving schools ads promising you can take the test on their cars(still not your own, but at least the one you got used to during training).

In Poland? Huh, that's news to me(but admittedly, it's been a while sinve I've taken my test :D )

Subaru CVTs already have essentially this feature. The transmission doesn't have gears (duh, it's continuously variable) but it still has the ability to manually shift to predefined ratios.

And it's horrible. I was a dedicated manual driver, but 3 years in an Outback with CVT and flappy paddles - it's miserable as an analogy of shifting. You can trigger several "down shifts", but there's some elastic feel in the system that doesn't convert to slowing the car via engine revs or give the feeling of immediate power. I gave up on the paddles early on.

Their CVTs also use these simulated gears during normal automatic driving, presumably just to make it feel more like a traditional automatic (which it mostly does, to my untrained senses). Apparently this is important to the car-buying market, although I don’t understand why.

Car reviewers love to complain about how a CVT lets the engine rev up to high RPMs and just stay there the whole time you're accelerating. Which is of course the entire point, but some people apparently prefer that their car spend lots of time at suboptimal RPMs, probably so that they don't have to break the habit of flooring it unnecessarily.

It's completely subjective but it just sounds wrong. Like being in a car with a slipping clutch. Apparently enough people feel the same the it's easier to fake real gears than change the market.

According to physics an ICE is most fuel efficient at 90% throttle and low RPM. At high RPM, or lower throttle positions you lose efficiency. So if you want maximum fuel efficiency you need to put the throttle to the floor, but stay at low RPMs. Thus the CVT is not doing the right thing for fuel efficiency. (though I'm not sure if the difference is worth measuring)

Obviously the exact numbers varies a bit from engine to engine.

When you put the pedal to the floor, you're not trying for efficiency, you're trying for power.

Dropping to lower RPM for efficiency is what happens when you're done accelerating and ease up on the pedal. That's not what people complain about with CVTs. The complaints are about behavior and sound during acceleration, not after acceleration when you are merely cruising on low power output.

That might be you, but not me. I'm always trying for effiency. My cars never have the acceleration to make power interesting. I'm weird, but whatever.

> nobody is doing much to make an electric car pretend to be something it's not

This is actually one of the aspects of the Leaf that most impressed/surprised me - Nissan put a lot of thought in to making it behave like an ICE car with an automatic/CV transmission.

For example, when the car is in D mode, sitting on flat ground, if you take your foot off the brake, it'll roll forward at low power. Similarly, taking your foot off the accelerator while rolling faster will apply just a little bit of regeneration.

The Prius has that too. It doesn't need to, but as you say, they wanted it to feel like a regular car.

It makes more sense to me to be honest. If I want to roll just a little bit to get a better position in a parking box, I don't want my foot to leave the break or be super careful about accelerating. I want to just barely move / not-be-stopped, and this implementation works just fine.

Yeah, my father-in-law drives an automated manual car which doesn't do this automatically. It's actually a strange car to drive because of that, it also changes at a snails pace so you have to drive pretty leasurely.

Our car which is the Ford Power shift dual clutch system which changes pretty much seamlessly and does creep on its own.

Don't all electric cars do this? I know all Teslas have this option (you can disable it in settings if you wish).

Electric motors have maximum torque from 0 RPM, which removes most of the need for a transmission. Think of it as a transmission being a patch for a limitation in a combustion engine. If you were switching from electric to ICE, you’d need to invent this patch.

And yet... for whatever torque the motor supplies at 0 RPM, you could have 4 times as much torque with a 4:1 1st gear.

As I see it, a manual stick EV would be pretty good in combination with a modestly-powerful drive system like what's in the Nissan Leaf, as opposed to something like a high-end Tesla (where you have enough power that shifting gears doesn't matter as much).

There's no need. We're very much EV-only since 2014, am yet to drive one that is underpowered in any practical driving circumstances.

> And yet... for whatever torque the motor supplies at 0 RPM, you could have 4 times as much torque with a 4:1 1st gear.

There's no point. The tyres can't provide that much traction. My Leaf wheelspins if I put my foot down at 0 RPM (until the traction control kicks in to moderate the torque). I assume it has no gearbox.

That's actually something I've wondered about. It seems like what you're saying may be true, but on the other hand you have cars like the White Zombie. (I know it's builder, John Wayland; he's giving me advice in my own more modest conversion.) The car started out as about a 1500 pound 70's Datsun, and with a dual-motor setup and some expensive batteries it can supposedly do 0-60 in about 1.7 seconds or so. If the Leaf is already at the point where adding more power doesn't help because you'd just spin the tires, how do cars like the White Zombie get around that?

The Leaf is front wheel drive I believe whereas the White Zombie is rear wheel drive. That probably accounts for some of it. Maybe the WZ also has a limited slip differential and some kind of high-traction tires? I don't know; drag racing is pretty far outside my area of expertise.

(Another random anecdote about the White Zombie: according to John, he always wanted to have a transmission, but they kept breaking whenever he tried to use one, so eventually he just settled on direct drive.)

Part of the issue is that the LEAF has tires that are terrible for anything approaching spirited driving. They are low rolling resistance round blocks of approximately cast iron and slip at the slightest provocation as a result.

(And yes, the LEAF is FWD-only.)

Pretty much all the early DC and golf cart battery based EV vehicle conversions used manual transmissions for exactly this reason.

sure. not sure how that's relevant to my comment explicitly talking about a car that doesn't have an actual transmission?

Yeah, me neither to be honest. I think I clicked on the wrong reply link.

Having maximum torque from 0RPM removes the need for a clutch, not a transmission. You still need to keep the motor within its effective RPM range, which is certainly easier for electric motors than ICE engines but is still a consideration. Many single-speed electric cars have much lower top speeds than their power output would otherwise allow.

I've always wanted to replace the gearstick in an AWD electric car with a torque split lever that controls both acceleration and braking distribution and basically gives you thrust vectoring. Pull it back under acceleration and you can spin the rear wheels, or under braking and you can do a handbrake turn. Push it left or right in a turn to control drift angle. When stationary, go full left or right to skid steer on the spot. The stunts you could do with such a setup would be insane.

Many EVs already work this way. Mine has 6 "speeds" controlled by paddle shifters that create "engine braking" by changing the regenerative braking level. It is 100% a software/UI thing because if I press the brake pedal moderately it does the exact same thing, but continuously instead of in discrete steps.

For anything that isn't a car you're gonna want gears for the same reason electric drills have gears. Think of it like low range.

"muh instant 0rpm torque" is absolutely nothing compared to the inertia of a hired driver who doesn't care (or is just a novice) dumping the clutch. Building a transmission to handle electric motor loads isn't a particularly remarkable engineering challenge. (Though you can make any engineering exercise challenging if you add sufficient weight and budgetary constraints)

Imagine a medium or heavier truck that's expected to both get a load moving at low speed and to cruise at 80. Putting anything near peak power through an electric motor at low speeds lets the magic smoke out in short order. Sure you can gear it down but then you have insane rpm at high speed and the wonderful engineering problem of getting that insanely overbuilt motor to not fling itself apart at those speed. A simple two speed transmission is a far more elegant solution. Especially since a transmission for an electric motor can be far simpler than one for an ICE. Letting the magic smoke out was a big problem in early EV conversions because power was low and consequently the vehicles available for conversion didn't have lots of good options for deep gearing.

I've done both. Clutch dumps require that you rev the engine up to high RPMs first. Instant torque is, well, instant.

I suggest you try it before you make fun of it.

Oh I know that a clutch dump is a far less elegant way to get moving than having torque available at 0rpm but the former is far, far harder on things.

The point I'm trying to make is that the people acting like electric motors are particularly hard on drive-line components are just wrong.

Avoiding overheat conditions is something the vehicle engineering needs to deal with. Teslas solve this problem with liquid cooling and the motor controllers use temperature sensors in the motor core to determine if the currents are too great for the cooling. So really, no issue. I expect the Tesla Semi will handle heavy loads at low speeds just fine. As for clutch dumping, I'll take electronic stability and traction control over the smell of burning clutch any day. Plus, the traction delivery is better in an EV; much less rotating mass means more rapid response and a better grip on the road.

And depending on how much performance "dealing with overheat" conditions necessitates leaving on the table a multi transmission might be their solution. Porsche is doing just that.

<i>For anything that isn't a car you're gonna want gears for the same reason electric drills have gears. Think of it like low range.</i>

Not with modern motor controls. For the last 20 years or so, most new locomotives have used 3-phase AC servo drive motors. It's striking to see one exerting enormous torque to get a train started, yet the Diesel engine generating the power is barely above idle. Also, no wheel spin. All the wheels can be synchronized electrically. This almost doubles the useful traction.

Early Teslas had a two-speed transmission. It was a flop. The shock loads on gear shifting were too large and the transmissions wore out rapidly. Tesla had to replace all of them. The big advance was liquid-cooled motors. Finally a way to get rid of the excess heat fast enough that you could use full power at zero speed if desired.

Locomotives have a very different relationship with weight than road going vehicles do.

Of course an economy car will have a single speed transmission but vehicles that need to put out peak power at both low speeds and high speed (e.g. the Taycan, and an F250) are going to benefit greatly from a way to optimize for multiple speed ranges.

I doubt it. There's no reason to bother doing the research to add another mechanical component which can wear and fail early, when you can just put a more powerful electric motor in.

That's the thing: the incentive structure has changed. The better use of time and research dollars is going to be just upgrading that motor till it's adequate. There's a logical cap on the maximum road speed you need to accomodate, so the problem is eventually "solved" in practical terms.

We have already put the effort into making transmissions though. Not a lot off research dollars needs to go into the problem at this point.

It might or might not be worth it, but transmissions that last as long as the car with no maintenance are common (oil changes are recommended, but few people do).

> instant torque that an electric vehicle puts out just shreds a gearbox

Doesn't the Porsche Taycan have 2 gears? and it actually beats Tesla Model S in 0-60 times

Plaid S is due this year to take back the crown. Curious if it adds a gear or just bigger motors.

Model model S (okay, any dual motor version) have different gearing on the front and rear wheels, so that more power can be sent to the front as speeds increase.

Roadster is due out this year with a 1.9s 0-60 (and an 8.9s quarter mile).

Can anyone here lend me a quarter million please?

It would be cool to have a stick-shift in the car that was fake, and would just make the car play different sounds as you "shifted" and allow you to control your playlist and volume levels with it...

Somebody put that in an early electric performance car. You actually had to shift, and got audio effects, but it was just software emulating gears. It was laughed at in the automotive press.

People don't like fake.

Sure we do, we just don’t like bad fakes.

(Bad is subjective, but the point remains)

Program it like a video game stick shift, let the software do the shifting

Don't diesel-electric locomotives use transformers for gear shifting? I assume if it's even needed for electric cars, they'd use something like that.

> instant torque that an electric vehicle puts out just shreds a gearbox.

On other hand, you can put way more beefier gears, as you need less of them.

I think top gear did this when they staged a drag race between the original Tesla and its Donor car.

The amount of torque and electric vehicle can put out is huge.

Porche Taycan has two speed transmission and Bosch has a turnkey 2 speed for EV applications. But seeing the numbers the Taycan puts down show it isn't a humongous difference like in an ICE. Probably not worth the cost/complexity as we see the Lucid Air can do qtr mile times in less than 10 seconds and soon upcoming Plaid Model S will have a possible quarter mile time of under 9 seconds and 200mph top speed.

is there some type of clutch that allows the rear wheels to free-spin at higher speeds?

On Tesla cars both motors are physically connected to the gearbox/axles with no clutch. They are doing careful tuning of the motor inverter controls to minimize energy use from motors that aren't under active load. Tesla refers to this as "Torque Sleep"

I imagine the process is basically tuning for the inverter output that minimizes power consumed as well as drag on the axle when the motor is not supplying active torque, or at least supplying less active torque than the other axle.

Offtopic but you seem to know a lot. I’ve been wondering do Tesla’s just have one regenerative braking force when you let off the gas?

Doesn’t this leave a lot on the table?

Wouldn’t you get more energy by having say gentle braking regen and harder braking regen?

You have analog control over the regen braking with the accelerator pedal. Push down to accelerate, lift up a little to maintain current speed, lift up a little more to start slowing down, lift off all the way for maximum regen.

It takes a week or two to unlearn the habit of "popping" your foot off the accelerator anytime you want to coast without necessarily slowing down, but after that it basically feels like your right foot is directly connected to the speedometer in both directions.

If you judge distances well you can still pop your foot. The deceleration curve is optimized to be smooth (minimize jerk?).

After a week you can judge pretty well. I use my brake pedal maybe once a week nowdays unless i'm driving alone and wanna feel some Gs.

Is it hard to switch back and forth between a normal car after that?

Hard? No. Irritating? Not surprisingly. I've had ICE loaners when my car is in for repair, and missing "one pedal driving" happens a lot, but it doesn't stop me from operating the ICE vehicle.

Perhaps weirdly the thing I think it makes the hardest is possibly videogames, which exaggerate controls in their own ways. I've gotten very frustrated with braking in most difficulty settings in the Forza series after driving an EV with "one pedal driving" capability for years now.

Brakes are easy compared to keys. I keep getting out of ICE cars and just leave them running. Whoa, you need to tell a car you got out? Why?

I own two very different motorcycles, a Ducati Monster900 (a 900cc twin four stroke) and a Cagiva Mito (a 125cc single cylinder 2 stroke).

The Mito, you can let go of the throttle, have it snap back to zero, and you barely slow down at all no matter what gear it's in. There's practically no engine braking at all. If you want to slow down you have to grab the brakes.

The Monster, in lower gears, will skid the back wheel if you chop the throttle from 75+% of peak rpm. You can pretty much get that bike down to walking pace in a fairly short distance without touching the brakes at all. (Thats a game I sometimes play on it, see if I can get home from work without touching the brakes, just slowing down using the throttle and gears.)

Don't know about the Teslas but the Porsche Taycan does exactly that but for energy saving.

Haldex type of 4 wheel drive do this, but in that case the rear wheels are disconnected at most times, they are engaged only when the car feels the need (hard acceleration, etc).

It's not a big deal. In EV cars it is more complicated as they usually have an engine per wheel so you'd need a clutch at each wheel.

Doesn’t the taycan have some kind of gears?

I think so.

Reply on news.ycombinator.com